Key Takeaways
- Factory is partnering with Databricks to bring autonomous software engineering into enterprise data and AI environments, with Droids building, migrating, and securing software on Databricks using existing data, models, and governance.
- A widely shared critique describes a "big dip" in developer curiosity, with learning increasingly outsourced to AI and engineering leaders repeating the same PR-approval, coding-automation, and AI cost/model-evaluation work.
- A no-code founder case study shows AbMaxx reaching roughly $55K/month and 3,000+ paid subscribers, with a rebuild in Rork and a TikTok/Meta ad playbook behind the growth.
- Noah Kagan used Claude to walk his own signup and onboarding flow, surfacing a missing core step, while Alex Lieberman offered a readiness-versus-impact framework for prioritizing AI use cases.
- A bootstrapper-versus-VC comparison argues profitable founders can retain better personal outcomes, and @levelsio warns sellers against accepting earnouts in acquisitions.
1. AI and Software Engineering
- Factory is working with Databricks to bring autonomous software engineering into enterprise data and AI settings; Droids can build, migrate, and secure software directly on Databricks while using the enterprise's existing data, models, and governance systems 1. The announcement frames this as applying autonomous engineering inside established enterprise controls rather than as a standalone coding tool.
- A developer-culture critique describes "the big dip": many developers are no longer curious about how code works internally and instead spend energy on games and other diversions, and while they can use AI, the author argues that is not enough to extract greater value from it 1. The same account relays an observation that developers have stopped reading technical books or taking courses, learning almost entirely through AI and even outsourcing full thought processes to it 2. Separately, a forwarded view says engineering leaders are all repeating the same work—automated PR approval, coding automation, AI cost optimization, and model switching/evaluation—with the comment that solving this would make someone "the new Oracle" 3.
- Noah Kagan recommends founders use Claude to simulate a new user going through their website signup and onboarding, asking it to list the five most important fixes and what it sees at each step; in his own test, the flow never asked users to create an appointment type, which he says is the product's core, and Claude then produced a sketch of a fix 1. Alex Lieberman proposes prioritizing AI use cases on two axes—readiness (can it be executed today) and impact (size of the return)—accelerating high-impact, ready work; investing to unlock high-impact work that is not yet ready; treating low-impact, ready work as quick wins; and deprioritizing low-impact, unready work 2.
2. Founder Case Study and Monetization
- Lino, 18, had never written code, left college after one semester, connected with app founders through X, and built AbMaxx, which scans abs and outputs body-fat percentage, a lower-ab score, and a training plan; he says the niche has no direct competitor 1. The app reportedly earned $7K in its first month and now runs at about $55K/month, with 16K downloads in 28 days and 3,000+ paid subscribers 2. He spent two months building a first version with prompts in Lovable, abandoned it, then rebuilt V2 in Rork in one month 3. On marketing, early random UGC spending lost thousands of dollars at 5%–10% margins; he later gave equity to a large fitness influencer, bringing ad cost near zero, scaled on Meta at 2–2.5x ROAS, then moved the same creative to TikTok for 4–5x ROAS among 15–30 year olds 4. His five-step relaunch playbook: enter the X app community; validate demand with Sensor Tower and faceless "comment APP"-style TikTok slideshows; launch in Rork within seven days; pay a micro-influencer $100 for ten videos to find a working hook; then scale UGC and paid ads 5. The account's takeaway is that the app itself is replicable and Lino's difference was willingness to launch, fail, and reorganize old marketing ideas 6.
- A forwarded argument compares a bootstrapped founder earning $500K/year with a founder who raised $20M: the latter is constrained by investor return expectations, may draw a $200K salary, and could hold only 20% after multiple rounds, while the former owns 100% and can take distributions, reinvest, or sell—a $4M acquisition can be meaningful, whereas the VC-backed founder may need a $50M exit for comparable personal proceeds after dilution and investor preference terms 1. The post acknowledges VC can help build large companies but argues attention overweights funding amounts and underweights what founders actually keep, urging more founders to take pride in profitability rather than rounds 2. On exits, @levelsio advises never accepting an earnout when selling a company and insists on cash, saying he has declined such arrangements multiple times 3. A quoted post cites Poppi's sale to PepsiCo at $1.9B plus a $300M earnout that Pepsi valued at only $12M, while stressing the author does not know the case details but that such large shortfalls often lead to litigation; it lists earnout risks—buyer control of budget, pricing, and hiring after closing; misaligned incentives; accounting definitions of revenue and EBITDA driving payouts—and recommends tying payouts to harder-to-manipulate metrics such as revenue, adding operating commitments, audit and dispute mechanisms, acceleration if the buyer sells or shuts the line, and shorter measurement periods, while valuing the deal on closing cash and treating the earnout as a bonus rather than money owed 4.
3. Distribution, Products, and Market Signals
- A forwarded view argues that people who say "distribution is the only moat left" have, in 99.3% of cases, never done any distribution beyond making that claim 1. Another forwarded point warns that the ability to build software that looks like it could serve a million users before understanding a single user's real need invites self-deception; instead, find people already doing the relevant work, give them early access, and watch them use the product in unexpected ways—the faster you build, the sooner you need contact with reality 2. A separate forwarded survey-style claim says 10,000 respondents said they should keep using Salesforce; if choosing again they would consider HubSpot, Attio, and Affinity for @saastrfund, but most importantly they would not leave or migrate and would stay on Salesforce—leading to the conclusion that winning customers from a competitor now requires persuading not just people but also agents 3.
- acquire.com listed several businesses for sale: a subscription platform for small-business owners in the health-economy space with $378.4K TTM revenue asking $1.1M 1; a profitable mobile fitness SaaS/app with $1.7K MRR and $25.4K TTM revenue asking $96.6K 2; a team using no-code form tools to collect and analyze data with $345K TTM revenue asking $624K 3; and a real-time AI sports research iPhone app with 300K+ downloads, 400 paid subscribers, 1K+ 4.7-star reviews, and $164.9K TTM revenue asking $413K 4. The account also says that before buyers see a listing, the numbers have been reviewed, expectations set, and the seller prepared for follow-up conversations 5. Separately, @levelsio forwarded a claim that games previously impossible because production cost far exceeded return will become viable, and games developers could not build due to resource limits will appear, predicting a new golden age for gaming 6; the same account also forwarded a Reception launch describing in-app customer support for iOS apps that claims to turn chats into sales, ask satisfied customers for reviews, reduce churn, persuade users during onboarding, and abandon chat at checkout 7.
- Patrick McKenzie says he had seen predictions that "adversarial contact points" would saturate but this is the first good chart showing the trend, and he expects the relevant authorities to take a long time to recognize what is happening; the quoted post charts applications, appeals, and complaints across 12 government agencies worldwide, with a gray dashed line marking when ChatGPT appeared 1. He also notes seeing Claude complain that GitHub keeps pursuing "nine nines" of availability, joking that his software has its own opinions and now opinions about other software, and may soon compete for favor like geek court politics 2.
